>> Panel: Multimedia applications: Beyond similarity searches

Researchers working on multimedia applications have spent much time on designing efficient low-level support for image similarity searches (via the many multidimensional indexing schemes), spatial joins, high-dimensional clustering, and high-level classification. Generally speaking, the common approach assumes some similarity metric used in a query by example paradigm.

From the multimedia users' point of view, many proposed applications need to integrate different media types with novel ways of querying, accessing, visualizing, and summarizing large collections of multimedia data. In these applications, the starting point of a query is not necessarily an example given by the user.

The goal of this panel is to present and review some of the needs of novel multimedia applications that go beyond similarity searches. This may help database researchers focus on providing the necessary support for these applications. <\p>

> Moderator: John Smith (IBM, T.J. Watson Research Center, USA)

> Panelist: Nalini Ratha (IBM, T.J. Watson Research Center, USA)

> Panelist: Amarnath Gupta (San Diego Supercomputer Center, University of California San Diego, USA)

> Panelist: David Doermann (University of Maryland, College Park, USA)

> Panelist: Jonathan Goldstein (Microsoft Research, USA)

> Panelist: Uri Shaft (Oracle, USA)